Home/ Part XIII — Expert Mode: Systems, Agents, and Automation/42. Fine-Tuning vs Prompting vs Retrieval (Decision Framework)/42.2 When fine-tuning hurts (rapidly changing knowledge)

42.2 When fine-tuning hurts (rapidly changing knowledge)

Overview and links for this section of the guide.

On this page

When It Hurts

Fine-tuning is wrong when:

  • Knowledge changes frequently (use RAG)
  • You need citations (fine-tuning can't cite sources)
  • Data is insufficient (<100 examples is usually too few)
  • General capabilities matter (fine-tuning can reduce them)

Problems

// Fine-tuned model for product info
Q: "What's the price of Product X?"
A: "$99" // Was correct in training data

// But price changed last week...
// Fine-tuned model: still says $99 (WRONG)
// RAG approach: fetches current price (CORRECT)

Where to go next